Strategic targeting: authoritarian capacity, state dependent populations, and electoral manipulation

ABSTRACT

Electoral manipulation is an activity conducted primarily at the local level, and the emerging comparative literature has shown that patterns of manipulation often exhibit striking spatial heterogeneity within countries. Yet, the causes of this variation are less well understood. This article explains variation in electoral manipulation at the local level, and presents a theoretical framework focusing on the role that local socioeconomic conditions play in affecting regime agents’ capacity to carry out electoral manipulation. Specifically, we argue that local-level heterogeneity in electoral manipulation can be explained by regime agents’ targeting of manipulation at localities where the electoral returns of illicit efforts are greatest and the risks associated with manipulation lowest. We test our hypotheses with multilevel statistical models that use unique local (county) level sociodemographic and electoral data from Russia. Our findings contribute to the comparative literature on electoral malpractice and authoritarian politics.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 In this paper, we refer to mobile voting as a “non-standard” voting practice, as it is a method that allows voters to vote outside their primary polling station on Election Day. Other established terms in the comparative literature such as “distance voting”, or “early voting” measures, do not encompass mobile voting practices.

2 T-tests indicate that these differences are highly statistically significant.

3 The dataset includes data drawn from the All-Russia National Census, the Central Electoral Commission of Russia and the Russian State Statistics Service (Rosstat). The data used in this article includes observations from all the 80 subnational regions where consistent socio-economic data is available from Rosstat.

4 Accessed at http://www.gks.ru/free_doc/new_site/perepis2010/croc/perepis_itogi1612.htm

5 Accessed at http://www.gks.ru/wps/wcm/connect/rosstat_main/rosstat/ru/

6 For a robustness check, we also ran our models using a dependent variable that was coded one (“1”) if the percent of mobile ballots cast in each rayon was greater than 1 standard deviation above the rayon-level national average. The results were substantively similar.

7 This denotes political parties that are in nominal opposition to the Kremlin but de facto vote in line with the regime in the Duma.

Additional information

Funding

This work was supported by the Finnish Cultural Foundation and the Academy ofFinland (Grant numbers 258190 and 319883).